CA1081816A - Bidirectional digital information transmission system using a single two-wire line - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE:

A bidirectional digital transmission system having two end stations interconnected by a single two-wire line along which are inserted repeaters having a controllable amplification di-rection. Each end station includes transmission equipment com-prising a transmission store unit in which incoming digital in-formation is written-in at a rate of d bits per second and in which the such written-in information is read-out at a rate higher than 2d bits per second under control of a transmission control unit. A time-division multiplexer multiplexes a syn-chronization word and other auxiliary information with the such read-out information onto said line. A generator produces an analog control signal for the amplification direction of the re-peaters. The multiplexed information and the control signal are mixed and transmitted onto the two-wire line. Each end station also includes receiving equipment comprising a receiving store unit and a receiving control unit which controls the receiving store unit for writing thereinto incoming digital information at a rate higher than 2d bits per second and for reading out the latter digital information at a rate of d bits per second.

Description

-- ~L0t318~6 The invention relates to a system for transmittingdigital information between two stations connected by a two-wire line, e.g. a coaxial line or a symmetrical balanced line.
In order to exchange coded information such as pulse code modulation (P.C.M~) signals or digital data between twostations, known transmission systems generally use a transmission channel such as a four-wire line, a two-wire channel being used for transmitting information in one direction and the other two-wire channel being used for transmitting information in the other direction. The cost of the wired cable containing the four-wire channels is an important element in the cost of a digital trans-mission system.
An object of the invention is to reduce by half the number of wires used for bilateral transmission of information, even at the expense of increasing the complexity of the end station and of the intermediate repeaters, if any, inserted along the transmission line if required owing to its length.
Some known bilateral transmission systems use two-wire transmission supports, e.g. in the "N+N" system, in which infor-mation is transmitted in the form of analog signals locatedin different frequency bands for each transmission direction. This method does not appear easy to adapt to the transmission of numerical signals.
It is also known from French Patent Spec. 2 244 315 dated 12 September 1974 to use two-wire lines for connecting telephone subscribersreceiving and transmittingP.C.M. signals to a subscriber concentrator connected to a digital exchange by a time-division multiplexing line. In this system, the concentrator is connected to the numerical exchange by a two-directional time-division 30 multiplexing channeltransmitting 32 octets in a 125u sframe, l.e.
at a flow rate or rhythm of 2.048 Megabitsper second. The octets received from the exchangearestored in areceivingstore in the concentrator, whereas the octets intended for a particular -- 10t318:16 subscriber's station are re-transmitted by the two-wire line of the subscriber's station at a lower rate, e.g. 256 kilobits/sec..
After receiving an octet, the subscriber's telephone set transmits an octet to the concentrator at a low rate, and the octet is stored in a transmission store in the concentrator and re-transmitted at a high rate to the exchange.
The preceding method cannot solve the general problem of transmitting digital information of any nature between two stations connected by a two-wire line. This is firstly because the method mainly relates to the network formed by the subscriber's lines connected to a concentrator, the network having a star structure in which the flow rate transmitted along each arm is much less than the rate of the main flow entering the concentrator, whereas the invention relates to a connection between two stations. Another reason is that the method applies only to digital information in a time-division multiplexing telephone network, which means that the duration of the multiplexer scanning frame (125 ~s) has to be taken as the duration of the full transmission cycle on the two-wire line.
According to the invention, there is provided a bidirec-tional digital transmission system between two end stations inter-connected by a single two-wire line along which regenerative repea-ters having a controllable direction of amplification are inserted, in which each end station comprises transmission equipment including a transmission store unit, means for writing incoming digital infor-mation into the transmission store unit at a rate of _ bits per second, means for reading out information written into the trans-mission store unit at a rate greater than 2d bits per second, means for transmitting the read out information to the line, and means for adding thereto a signal controlling the amplification direction of the regenerative repeaters, and in which each end station further comprises receiving equipment including a receiving store unit, means for writing incoming digital information into the ". ,~ ,.
,~ .

.
. .

receiving store unit at a rate greater than 2_ bits per second, and means for reading out the digital information written in the recei-ving store unit at a rate of _ bits per second.
According to another feature of the invention, the trans-mission system also comprises means for selecting one or the other direction of transmission in the end stations and, if required, in the repeaters and/or regenerative repeaters inserted in the two-wire line, the said means comprising switching means and transmit-ters and receivers of one or more special signaIs having the same nature as the signal to be transmitted or having a different nature, i.e. either one or more digital or one or more analog control si-gnals, the transmission system also comprising means in the end stations for controlling the selection means according to a periodic sequence.
According to another feature, the means for controlling the means for selecting the transmission direction comprise a clock controlling a transmission logic circuit and a receiving logic cir-cuit in each end station. The control means are autonomous in one of the end stations, which is considered as the control station, whereas in the other end station, which is considered as the con-trolled station, they are themselves controlled by the signals received from the control station.
According to another feature of the invention, interfaces are inserted between the stations and the general network and com-prise stores in which the information bits coming from or en route to the general network are written in or read out continuously at a rate of d bits per second, and the information bits going to or coming from the two-wire line of the device of the invention are discontinuously read out or written in at a rate greater than 2d bits per second.
According to still another feature, the writing-in and reading-out operations of the stores forming the interfaces bet-ween the system according to the invention and the general .
' -network are controlled by ~he logic assembly of the end stations of the device according to the invention.
Other features of the digital transmission system according _ to the invention will be clear from the following description, which is illustrated by the accompanying drawings, in which :
Fig. 1 is a diagram of the bilateral transmission system according to the invention, comprising a two-wire line;
Fig. 2 is a time diagram of transmission, when the duration of transmission is the same for the two stations;
Fig. 3 is a similar diagram, when the duration of transmission is different for the two stations;
Fig. 4 shows an end station in the digital transmission system;
Fig. 5 shows a regenerative repeater inserted in the two-wire line; and Figs. 6 and 7 show two ways of arranging the stores disposed in the end stations.
Referring firstly to Fig. 1, references 1 and 2 denote two stations A and B connected by a two-wire line 3. Regenerative repeaters 5 are inserted in line 3.
Starting from a time to (Fig. 2), station A transmits digital information for a period of I seconds to station B via transmission line 3 operating in the direction from 1 to 2. The numerical information arrives at 2 between (to + ~) and (to +T +~). At the end of the holding time tgb~ i.e. at the time (to +T +~ + tgb) station B transmits digital information for a period of T seconds to station A via transmission line 3 operating in the direction from 2 to 1, the digital information arriving at 1 between (to +T + 2~ + tgb) and (to + 2T + 2~ + tgb). After a holding time tga, station 1 can again transmit to 2, and so on. The times at which transmission from 1 to 2 starts are :
to + iT = to + i( 2T + 2~ + tga + tgb) and the times at which transmission from 2 to 1 starts are :

, . . . , .: . . . . . , . . . , . .
:- . .- , .
: . . . . .
~ .
.. . . . . . .

1~81816 to + T + ~ + tgb + iT = to ~ r + ~ + tgb ~ i(2T + 20 + tga + tgb) If tga = tgb = tg, the times at which transmission begins are for 1 to 2 :
to + 2i ( T+~+tg) and, for 2 to 1 :
to + t2i+1)(T+~+tg) If the flow rates in the two directions are different (Fig. 3), the transmission periods are Ta from l to 2 and Ib from 2 to 1 and the period T becomes :
la + Tb + 2~ + tga + tgb It can be seen that, if the propagation time and the holding time are neglected, T = 2~. Consequently, the transmission rate of data on the two-wire lines must be at least 2d, i.e. at least twice the transmission rate d of date on the four-wire line. Owing to the propagation time, the holding time and the fact that synchro-nization bits and auxiliary bits can be inserted, the "two-wire"
rate must be more than twice the "four-wire" rate, i.e. equal to (2d + ).
Referring now to Fig. 4, 11, 12 denote the four-wire line connected to the end station on the side of the general network, the two-wire line 11 being the incoming line and the two-wire line 12 being the outgoing line. A two-wire line 3 connects the end station shown to another end station in the system. Line 11 is connected to an inverting circuit 14 having two outputs connected to the inputs of two transmission stores 15, 16 respectively. The outputs of stores 15, 16 are connected to the two inputs of an inverting circuit 17, and the output of circuit 17 is connected to a time-division multiplexer 18. The multiplexer i5 adapted to add a synchronization word to the digital information to be transmitted, the synchronization word being preceded if required by a preamble and followed by various auxiliary information depending on the particular application, e.g. service-channel bits, ~81816 tele-control bits, tele-signalling bits or quality control bits.
It is advantageous to use this facility of inserting supplemen-tary bits in order to give the group a predetermined length in spite of the nature (usually plesiochronic) of the digital infor-mation sources, by adding stuffing bits and a word indicating the variable length of the useful part of the group; this faci-lity is show in Fig. 4. A transmission control unit 20 controls -inverters 14 and 17, writing-in and reading-out of transmission stores 15 and 16, and the insertion of the supplementary bits of the synchronization word, of themessage-length word and of the stuffing bits.
The digital train formed in multiplexer 18 is sent to a coder 19 which converts the code used in the station to the code used on the line. An analog control signal produced by generator 21 is added to the digital train and the resulting composite signal is amplified by amplifier 22 and applied to the two-wire line 3.
In the reception direction, the two-wire line 3 is connect~
ed to a bandpass filter 31 which transmits the control signal.
The filter is followed by a detector 32, and the detected signal is applied to a transmission control unit 20 and a reception control unit 40. Line 3 is also connected to a circuit comprising a locking amplifier 33, an equalizer 34, a variable-gain ampli-fier 35, a band cut-off filter 36 cutting off the frequency of the control signal, a regenerator 37 which also decodes in inverse manner to coder 19, a demultiplexer 38 and an inverting circuit 39. Demultiplexer 38 is used for extracting the synchronization bits and the stuffing indication bits from the digital train and applying them to the reception control unit 40.
The two outputs of circuit 39 are connected to two recep-tion stores 41, 42. The outputs of stores 41, 42 are connected to the inputs of the inverting circuit 43, whose output is ~ -6-.
- .
. , . ' .

3LC~81816 connected to the outgoing two-wire line 12. The reception control unit 40 controls inverters 39 and 43, writing-in and eading-out cf the receltlon .

,~, , .
., :
,,,/~

. ,~ .

. /

., . . _, -6a-8~816 stores 41, 42 and demultiplexer 38.
The entire end station is controlled by a clock 10. Since stores 41 and 42 are written in discontinuous manner and since reading-out must be continuous, it is controlled by a rhythm smoothing device 44.
Fig. 5 shows a regenerative repeater. On each side of the repeater, the two-wire line 3 is connected to blocking amplifi.ers 51 and 61 respectively, to bandpass filters 52 and 62 respectively and to amplifiers 53 and 63 respectively. Filters 52, 62 are respectively adjusted to the frequencies Fl, F2 of the control signals transmitted by stations 1 and 2. They are connected to detectors 54, 64 respectively, which in turn are connected to a logic control unit 50. Unit 50 controls the alternate blocking and unblocking of amplifiers 51, 61, inverters 58, 68 and control signal regenerators 59, 69, which will now be described.
Amplifiers 51, 61 are each connected to a circuit respectively comprising control cut-off filters 55 and 65, equalizers 56 and 66 and variable-gain amplifiers 57 and 67. The outputs of amplifiers 57, 67 are connected to the two inputs of an inverter 58 and the inputs of amplifie7s 53 and 63 are connected to the two outputs A of an inverter ~. A regenerator 60 is disposed between inverters 58 and 68. ~
A regenerator~for the control signal at frequency Fl is inserted between filter 52 and amplifier 53 ànd a regenerator~of the control signal F2 is inserted between filter 62 and amplifier 63.
It can easily be seen that, when the regenerative repeater 5 receives the control signal Fl, the control unit 50 unblocks amplifier 51, blocks amplifier 61, places regenerator 60 between the output of amplifier 57 a,nd the input of amplifier 53, unblocks ~ege~O~
the control-signal ~e~e~r~=~ 59 and blocks the control-signal regenerator 69.
When repeater 5 receives the control signal F2, the control .: .

- ~L0~3~816 unit 50 unbloc~s amplifier 61, blocks amplifier 51, places regenerator 60 between the output of amplifier 67 and the input of amplifier 63, unblocks the regenerator of control signal 69 and blocks the regenerator of control signal 59.
If ~he system according to the invention is applied to high-speed transmission, series~parallel and paxallel-series conversions have to be made at the store inputs and outputs respectively, so as to use store units such as MOS or CCD stores operating at speeds less than the transmission speed. If do is the minimum write-in or read-out rate of the store units, the required number k of units :in parallel must be at least (2d+E)/d where (2d+F) is the flow rate along the line.
Advantageously, the following use may be made of the need for series-parallel conversion in order to use high-capacity stores operating at reduced speed. Instead of a digital train at d bits/s, it may be preferable to have k digital trains at the rate of (d/k)bits per second at the transmitting station since, during high-speed digital transmission, the digital train is usually obtained by multiplexing a number of low-speed plesiochronic digital trains. For example, a digital train at 34.368 Mb/s is obtained by multiplexing 16 digital trains at 2.048 Mb/s.
Owing to the need to use store units in parallel, it is unnecessary to dispose a multiplexer upstream of the transmission end station and a demultiplexer downstream of the receiving end station in order to make the transmission from 2 to 34 Mb/s and 34 to 2 Mb/s respectively. Thus, the multiplexer operation is incorporated with store operation.
Two possible methods of arranging stores 15, 16 and 41, 42 are described with reference to Figs. 6 and 7. In both cases, it is assumed that there are 16 incoming digital channels and 16 outgoing digital channels at 2.048 Mb/s, i.e. an incoming or outgoing flow rate of the order of 34 Mb/s allowing for stuffing, `- ~L081816 and that the flow rate on the two-wire line 3 is of the order of 72 Mb/s and the groups contain 40,000 bits per incoming channel, i.e. a total of about 640,000.
In Fig. 6, 16 incoming channels 700 to 715 are applied to input inverters 740 to 7415 which send 16 groups, each of 40,000 bits, to the 16 store parts 750 ~ 7515. Writing-in the store occurs at the rate of 2 Mb/s. Read-out occurs at the rate of 4.2 Mb/s.
The 16 outputs of output inverters 770 ~ 7715 are connected to the inputs of a multiplexer 78 operating bit by bit. The output of mul-tiplexer 78 is connected to the input of multiplexer 18 in Fig. 4.
It can be seen that the store read-out rate remains acceptable, since the high rate along the two-wire line 3 results from the ac-tion of multiplexer 78 which interlaces the 16 channels in a single gro~.
In a manner analogous to the two stores 15, 16 in Fig. 4, the two stores 75, 76 shown in Fig. 6 operate in opposition, one store being written in when the other is read out. The switching period is of the order of 20 milliseconds.
In Fig. 7, the bits from the 16 incoming channels 800 ~
8015 are not interlaced into a single group as before, but distribu-ted among 16 successive sub-groups of 40,000 bits, i.e. one sub-group per channel. Multiplexer 88 disposes the sub-groups in series, and they are successively delivered by multiplexers 890 ~ 8915 at an output rate of about 72 Mb/s. Owing to the high rate, high-order multiplexing of the stores is necessary so that each fraction of a store can be read at a speed compatible with its technology. In Fig. 7 stores 85 and 86, which are alternately written-in and read-out, are each divided into 16 stores 850 to 8515 and 860 to 8615 associated with one channel. Each channel store is subdivided into 40 parts, e-g. 850 is made of 8500, 850 1 -- 850 39- It is preceded by a demultiplexer 83 having 40 outputs and followed by a multiplexer 89 having 40 inputs. The digital train on channel 800 is written in store 850 as follows: The first bit following a , command from the transmission control unit 20 is written in 8500, the second in 850 1' the third in 850 2 ~ the fortieth in 850 39' -' the forty-first in 8500, the forty-second in 8501 and so on up to about the 40,000th bit, and control unit 20 changes over store 85 to the read-out phase and store 86 to the write-in phase.
During the write-in phase of 85, all the channels are simultane-ously written in the respective stores, i.e. channel 800 in 85 channel 801 in 851, ... channel 8015 in 8515. During the read-out phase of 85, the stores for channels 850 to 8515 are succes-sively read, beginning with 850. The bits are read in the orderin which they were written and are placed in series bit by bit at 72 Mb/s by multiplexer 890 in the case of the bits read in 850, then by multiplexer 891 in the case of the bits read in 851,..., ,and by 8915 in the case of the bits read in 8515.
The advantage of dividing the group sent from one station to another into a number of sub-groups equal to the number of ; incoming digital trains is that each sub-group can be made inde-pendent and can be provided by multiplexer 18 (Fig. 4) with a preamble, a synchronization word, and specific auxiliary bits such as the departure address and the arrival address, so that the sub-groups are,independent of the other sub-groups and intermediate,stations can be produced, with insertion and extrac-tion of 2 Mb/s channels at the level of the sub-groups which represent them.
The two store arrangements which have been described with reference to Figs. 4, 6 and 7 differ with regard to the manner in which the bits from the various incoming channels are distributed (i.e.,dispersed or in groups) in the transmitted group. Figs. 4, 6, 7 have a common feature, in that two-stores equal capacity are used alternately for writing-in and reading-out.

" Other known methods of using stores may also be used ac-cording to the invention. More particularly appreciable reductions in store capacity and in the transmission time of the system can be ~Q81~316 obtained by using a single store per 2 ~lb/s track, divided into secto'rs which are written-in and read-out in sequence, a small-capacity margin being left so as to prevent any sector from being simultaneously written-in and read-out.

, . ., . . . . . - - -

Claims (3)

W??? IS CLAIMED IS :

1.- A bidirectional digital transmission system between two end stations interconnected by a single two-wire line along which regenerative repeaters having a controllable direction of amplification are inserted, in which each end station comprises transmission equipment including a transmission store unit, means for writing incoming digital information into said transmission store unit at a rate of d bits per second, means for reading out information written into the transmission store unit at a rate greater than 2d bits per second, means for transmitting said read out information to said line, and means for adding thereto a signal controlling the amplification direction of said regenerative repeaters, and in which each end station further comprises receiving equipment including a receiving store unit, means for writing incoming digital information into said receiving store unit at a rate greater than 2d bits per second, and means for reading out said digital information written in said receiving store unit at a rate of d bits per second.

2.- A bidirectional digital transmission system between two end stations connected by a single two-wire line according to Claim 1, in which said transmission store unit and receiving store unit each comprise two alternately operating stores.

3.- A bidirectional digital transmission system as claimed in Claim 1, in which the end stations are connected to an external digital transmission network by a number k of four-wire digital channels comprising k incoming two-wire channels and k outgoing two-wire channels, and in which the transmission equipment of each end station comprises means for writting incoming digital information in different locations in the transmission store unit respectively associated with incoming channels, the writing speed being equal to the flow rate D of the incoming channels, so that the flow rate of incoming information is d = kD; means in each of said locations in said transmission store unit for reading the digital information written therein at a flow rate greater than 2D bits per second, and a multiplexer of the K trains of thus read digital information so that the flow rate of information leaving via the two-wire line be greater than 2kD.